WMAP 
Wilkinson Microwave Anisotropy ProbeThe data made available through this page has been updated. The most recent version of this data may be accessed through /product/map/current/ Analysis of the WMAP data involved the computation of cosmological parameters assuming a variety of models and including a variety of datasets in addition to the WMAP dataset. The various model/dataset combinations that were tested are displayed in a matrix; the combinations that were analyzed have links in the corresponding cells to individual display pages that show the cosmological parameters computed using that combination of model and dataset. A printable version of each list of parameters is made available as a Postscript file through a link its corresponding the individual display page. Mnemonics are used to identify the models and datasets in the table and the individual model/dataset displays. The following table translate the mnemonics:
^{1} RECFAST is the code included in the Code for Anisotropies in the Microwave Background (CAMB) that calculates the fractional ionization of the light elements in the universe for a range of redshifts around z=1000 for the purpose of determining the cosmic microwave background power spectra. As mentioned previously, each model/dataset combination has an individual display page that shows the WMAP measurements of the cosmological parameters, with a link to a Postscript file showing the same parameters. These display pages also contain links to files containing WMAP's best fit C_{l}'s, best fit matter power spectra, and Monte Carlo Markov Chains. The Markov Chain are supplied in a compressed tarball that contains a README file describing the contents. The spectra files contain:
Some of the Markov Chains are flagged as "postprocessed". These chains have been constructed from chains originally run with a different data combination. Importance sampling (Lewis & Bridle, 2002) was used to reweight the original chain for the new dataset. The new distribution is checked against the original chain to determine if the statistics can be trusted. Some of the contents of these chains differ from normal chains:
To find the log likelihood of this chain, you should therefore use total lnlike = 'lnlike' + 'delta_lnlike' Additional Information
